jobyssey/apps/worker/schema.sql

-- D1 schema for the jobsdb snapshot (documents/postings/jobs.db). Applied first by
-- scripts/import-postings.ts as .import/0000-schema.sql; idempotent.
--
-- `postings` mirrors the source table column-for-column (all TEXT). `companies`
-- carries only what src/portals/jobsdb.ts reads. `postings_fts` is a
-- self-contained FTS5 table (not external-content: those need the shadow
-- tables, which `wrangler d1 execute --file` cannot bulk-load sanely), joined
-- on posting_id rather than rowid.

CREATE TABLE IF NOT EXISTS postings (
  posting_id TEXT PRIMARY KEY,
  company_id TEXT,
  company_name TEXT,
  company_domain TEXT,
  title TEXT,
  department TEXT,
  function TEXT,
  seniority TEXT,
  posted_date TEXT,
  location_text TEXT,
  remote_policy_stated TEXT,
  location_class TEXT,
  location_country TEXT,
  salary_min TEXT,
  salary_max TEXT,
  salary_currency TEXT,
  url TEXT,
  source_ats TEXT,
  source_tier TEXT,
  scrape_timestamp TEXT,
  classification_rationale TEXT
);

CREATE TABLE IF NOT EXISTS companies (
  company_id TEXT PRIMARY KEY,
  name TEXT,
  target_sector TEXT,
  crypto_native TEXT,
  priority_tier INTEGER,
  homepage_domain TEXT
);

CREATE INDEX IF NOT EXISTS postings_company_id ON postings(company_id);
CREATE INDEX IF NOT EXISTS postings_location_class ON postings(location_class);
CREATE INDEX IF NOT EXISTS postings_seniority ON postings(seniority);

CREATE VIRTUAL TABLE IF NOT EXISTS postings_fts USING fts5(posting_id UNINDEXED, title, company_name, location_text);